RESPONSE STRATEGIES IN AN INTERNATIONAL STRATEGIC
ALLIANCE EXPERIMENTAL CONTEXT:
CROSS-COUNTRY DIFFERENCES
Brian V. Tjemkes
VU University Amsterdam, the Netherlands
Olivier Furrer
Radboud University Nijmegen, the Netherlands
Koen Adolfs
VU University Amsterdam, the Netherlands
Arzu Ülgen Aydinlik
Istanbul University, Turkey
Journal of International Management
in Press
Correspondence to:
Olivier Furrer
Radboud University Nijmegen
Nijmegen School of Management
Thomas van Aquinostraat 1
P.O. Box 9108
6500 HK Nijmegen, the Netherlands
Phone: 0031 24 361 30 79
Fax: 0031 24 361 19 33
e-mail: [email protected]
1
RESPONSE STRATEGIES IN AN INTERNATIONAL STRATEGIC
ALLIANCE EXPERIMENTAL CONTEXT: CROSS-COUNTRY
DIFFERENCES
Abstract
International strategic alliances have grown increasingly popular in recent decades, yet their
failure rate is extremely high. Poor management of adverse situations contributes significantly to
such high failure rates. Moreover, the international environments in which international strategic
alliances operate exacerbate the adverse situations and make their management more critical.
However, extant research does not specify how people from different national cultures respond to
these adverse situations. In order to better understand cross-national differences, this study
investigates future managers’ preferences for specific response strategies in an international
strategic alliance experimental context. Using a scenario-based experiment with 1,379 business
students in five countries—Japan, the Netherlands, Switzerland, Turkey, and the United
Kingdom—the authors assess whether preferences for seven response strategies—exit,
opportunism, aggressive voice, creative voice, considerate voice, patience, and neglect—vary
across countries. The results indicate that national culture, both directly and interactively through
relationship-level exchange variables that characterize the adversity of the situation, influences
response strategy preference. This study advances literature on response strategies by explaining
that when faced with the same adverse situation, future managers from different countries likely
prefer different response strategies, depending on which response strategies they believe are most
adequate in their cultural environment.
Keywords: International strategic alliances, national culture, response strategies, exchange
2
variables.
3
1. Introduction
International strategic alliances (ISAs) are voluntary, long-term, contractual, cross-border
relationships between two firms, designed to achieve specific objectives through collaboration
(Brouthers and Bamossy, 2006). Because of their ability to capitalize on cross-border
opportunities, ISAs have grown increasingly popular, yet their failure rate remains high (Park and
Ungson, 2001). One of the key drivers of such failures is a lack of sensitivity to cultural
differences while managing adverse situations (Jiang et al., 2008; Meschi, 1997). When faced
with an adverse situation (e.g., low economic performance, poor relationship quality), managers
in an ISA must understand how national culture influences their partners’ responses to adversity
so they can reduce the likelihood of their alliances’ failure. Because ISAs take place within
unique socio-cultural contexts, managers also must consider normative beliefs about how people
should behave during interactions, which differ across countries (Doney et al., 1998; Thomas and
Au, 2002). Surprisingly then, despite the popularity of ISAs and reports of their high failure rate,
we have a limited understanding of how national culture influences people’s responses to adverse
situations (Aharoni and Brock, 2010; Nakos and Brouthers, 2008).
In an ISA’s context, previous studies have examined the influence of national cultural
differences between partners on relationship development (Barkema and Vermeulen, 1997;
Brouthers and Bamossy, 2006; Meschi, 1997), but they have not systematically identified how
people respond to adverse situations nor how culture influences their preference for certain
responses. For example, Barkema and Vermeulen (1997) demonstrate that cultural differences
between partners increase the likelihood of alliance termination, because differences breed
disagreement and conflicts. However, they only focus on alliance dissolution, without examining
other possible responses aimed at avoiding dissolution. Brouthers and Bamossy (2006) show that
4
within the context of Eastern–Western European joint ventures, cultural differences exacerbate
the detrimental effect of adverse situations, but their results indicate that managing adverse
situations and cultural differences fosters relationship development and reduces the likelihood of
alliance failure. However, their study does not detail the different response strategies that might
help manage adverse situations. To better understand how adverse situations can be managed in
international strategic alliances, we require more insight into the relationship between national
culture and response strategies.
To address this concern, we draw on response strategy literature in which response strategies
refer to the unique sets of responses that one party in a relationship uses to deal with adverse
situations (Hagedoorn et al., 1999; Ping, 1993; Tjemkes and Furrer, 2010). Extant response
strategy literature has focused on the effect of relationship-level exchange variables (i.e., distinct
types of adversity), such as partners’ satisfaction and exit barriers, on preferences for response
strategies (Geyskens and Steenkamp, 2000; Hibbard et al., 2001; Ping, 1993; Tjemkes and Furrer,
2010; Zhou and George, 2001). However, these studies usually implicitly assume that in a given
situation, people from any country respond the same way (Boyacigiller and Adler, 1991). Thus,
neglecting the issue that national culture is also likely to influence perceptions of the relative
importance of exchange variables (Thomas and Au, 2002; Thomas and Pekerti, 2003).
Thus, we develop and test a set of hypotheses that specify direct and interaction effects
among national culture (i.e., country), the relationship-level exchange variables that characterize
adverse situations, and response strategies. That is, we first propose that national culture directly
affects response strategy preference. Second, we examine national culture’s moderating effect,
through economic and social satisfaction, alliance-specific investments, and alternative
availability. The empirical results of a scenario-based experiment with business students (i.e.,
future managers) from five countries, namely, Japan, the Netherlands, Switzerland, Turkey, and
5
the United Kingdom, demonstrate that national culture not only directly influences response
strategy preferences but also increases sensitivity to particular types of adverse situations, thus
moderating the effect of the situation on preferences for certain response strategies.
We use a scenario-based experimental design and business students to test our hypotheses for
several reasons: Given the broad variety of ISAs that exists, both in terms of types and
governance forms, an experimental design allows us to control for extraneous factors and
confounding effects (Croson et al., 2007; Joardar et al., 2007). In addition, as our hypotheses
pertain to fundamental processes concerned with basic and relative stable characteristics of
human nature (Bello et al., 2009), which are not specific to ISAs but to more general
interpersonal relationships, business students who represent future managers are a pertinent
population to study response strategies in a strategic alliance context. Furthermore, the use of a
homogeneous sample of business students allows us to improve internal validity (Croson et al.,
2007). Moreover, business students are also more likely to only respond to the manipulations
than actual managers, whose responses may be contaminated by past experience with a specific
existing alliance (Bateman and Zeithaml 1989). The use of an experimental design with business
students, however, might raise the issue of external validity. To alleviate this issue, we test two
important assumptions of our experimental design by conducting a survey of actual alliance
managers. This survey allows us to assess the validity of our measures of response strategies and
the direct effects relationship-level exchange variables that are manipulated in the experiment.
The results show that the response strategy measures are valid and that the effects of the
exchange variables are as expected.
Accordingly we advance response strategy literature by demonstrating that national culture
influences people’ perceptions of the adversity of the situation, with an accordant interactive
effect on their response strategy preferences. By shifting the focus of our research from assessing
6
the overall effect of cultural differences on the adverse situation to the underlying direct and
moderating mechanisms by which national culture influences response preferences, our study
demonstrates that when faced with the same adverse situation, respondents from different
countries are likely to prefer different response strategies, depending on which strategies they
believe are most adequate in their cultural environment. Even if our results are obtained from
business students in an experimental context, they concern fundamental processes that are likely
to be present in actual ISAs. Therefore, the recognition of these cultural effects is useful to help
international strategic alliance managers, because with a better understanding of the behavioral
intentions of their foreign partners, as well as their own preferences, they can better overcome
adverse situations.
We organize the remainder of this article as follows: In the next section, we define and
classify response strategies. Then we introduce cultural dimensions to develop hypotheses about
their direct and moderating effects on preferences for response strategies. In the method section,
we describe the sample and the design of the scenario experiment we use to test the hypotheses.
Finally, we present the results and conclude with a discussion of the theoretical and managerial
implications of our study, along with limitations and directions for further research.
2. Theoretical Background
2.1. Response strategies
Hirschman’s (1970) exit, voice, and loyalty framework provides the foundation for an
important stream of research on response strategies. Hirschman initially represented exit, voice,
and loyalty as three alternative strategies along a constructive–destructive spectrum (Leck and
Saunders, 1992). Extending Hirschman’s framework with a fourth strategy, Farrell (1983) and
Rusbult and Zembrodt (1983) propose the ELVN (exit–voice–loyalty–neglect) typology, which
7
represents a parsimonious conceptualization of response strategies and derives its strength from
the underlying two-dimensional structure into which the four response strategies are organized:
an active–passive and a constructive–destructive dimension. In later studies, three types of voice
have been distinguished: considerate, creative, and aggressive voice (Hagedoorn et al., 1999;
Zhou and George, 2001). Opportunism also joins the typology as an active–destructive strategy
(Ping, 1993). Thus, seven response strategies appear in previous literature: exit, opportunism,
aggressive voice, creative voice, considerate voice, patience, and neglect (Tjemkes and Furrer,
2010).
Exit indicates a disinclination to continue the current relationship (Hirschman, 1970), which
is the ultimate and most destructive response to an adverse situation. It can be either active or
passive (Hagedoorn et al., 1999). By exiting the alliance, partners dissolve their relationship and
must find alternative ways to achieve their objectives (Rusbult et al., 1982; Withey and Cooper,
1989). Opportunism represents an active–destructive response (Wathne and Heide, 2000),
because it is an active attempt to increase benefits from the alliance in ways that are explicitly or
implicitly prohibited (Ping, 1993). Three types of voice represent three additional strategies:
aggressive, creative, and considerate (Hagedoorn et al., 1999; Tjemkes and Furrer, 2010; Zhou
and George, 2001). Considerate voice is constructive and slightly active and represents an
attempt to change, rather than escape from, the adverse situation by communicating in a
relationship-preserving manner and cooperatively discussing problems (Hagedoorn et al., 1999;
Ping, 1993). Aggressive voice is more destructive and more active; it refers to the forceful
imposition of views on alliance partners, without trying to avoid conflicts (Hagedoorn et al.,
1999; Hibbard et al., 2001). The third type of voice, creative voice, pertains to the generation of
novel and potentially useful solutions to address the situation, which makes it both active and
constructive (Zhou and George, 2001). With patience, a partner silently abides the issues, with
8
the confidence that things will improve in the future (Hibbard et al., 2001; Ping, 1993). It requires
voluntarily ignoring the issue, in the hope that the adverse situation resolves itself, and the belief
that undesirable circumstances are transitory phenomena that dissipate over time (Ping, 1993). It
thus is constructive and passive. Finally, neglect is passive but destructive, because it allows a
relationship to deteriorate (Rusbult et al., 1982). A neglectful manager believes that the alliance
does not deserve to be salvaged and expends little effort to keep it afloat (Ping, 1993; Pressey and
Qu, 2007).
2.2. Adverse situations and exchange variables
Building on social exchange theory (Blau, 1964) and interdependence theory (Thibaut and
Kelley, 1959), Rusbult and colleagues (Rusbult and Farrell, 1983; Rusbult et al., 1988) propose
that preferences for active–passive and constructive–destructive response strategies depend on
relationship-level exchange variables that characterize the nature of the adverse situation. Over
time, their investment model has been tested empirically in several contexts, including job
satisfaction (Rusbult and Farrell, 1983; Rusbult et al., 1988), romantic relationships (Rusbult and
Zembrodt, 1983), channel relationships (Geyskens and Steenkamp, 2000; Ping, 1993), and
strategic alliances (Tjemkes and Furrer, 2010). Among these studies those focusing on alliances
identified four exchange variables that influence preferences for response strategies: economic
satisfaction, social satisfaction, alternative attractiveness, and alliance-specific investments.
These exchange variables represent different adverse situations in ISAs to which managers must
respond.
Economic satisfaction pertains to managers’ evaluation of the financial outcomes of an
alliance (Geyskens and Steenkamp, 2000). According to Geyskens and Steenkamp (2000), an
economically satisfied manager considers the alliance a success with respect to goal attainment,
9
effectiveness, productivity, and the resulting financial outcomes. Prior response strategy research
has produced results indicating that economic satisfaction influences managers’ response
preference on the active–passive dimension but not on the constructive–destructive one. Low
economic satisfaction implies that managers perceive a discrepancy between prior expectations
and desired financial results (Geyskens and Steenkamp, 2000), which requires an active response
to improve the situation rapidly (Das, 2006; Tjemkes and Furrer, 2010). However, this active
response could be destructive, such as acting opportunistically to extract additional financial
benefits, or constructive, such as using creative voice to find new ways to solve the situation.
Regardless of how they do it, managers are more likely to “rock the boat” to restore performance
and increase their economic satisfaction, instead of waiting patiently for the situation to improve
(Ping, 1993). In contrast, managers who are satisfied with the economic performance of the
alliance likely behave passively (Geyskens and Steenkamp, 2000; Tjemkes and Furrer, 2010);
they can either be patient or neglect the issue.
Social satisfaction pertains to managers’ evaluations of the psycho-social aspects of an
alliance; it implies that interactions with counterparts are fulfilling, gratifying, and facile
(Geyskens and Steenkamp, 2000). Managers’ perceptions of relational quality affect their social
satisfaction; if relational quality is poor, the alliance suffers dysfunctional conflicts, distrust, and
low commitment (Ariño et al., 2001). Empirical results suggest that social satisfaction in turn
influences the constructive–destructive dimension but not the active–passive one. Low social
satisfaction creates greater suspicion about a counterpart’s intentions and reduces expectations
about the potential future benefits of the relationship (Geyskens and Steenkamp, 2000).
Therefore, managers dissatisfied with relationship quality may terminate the relationship rather
than try to save it through constructive responses (Tjemkes and Furrer, 2010). To do so, they can
respond destructively, either in an active way by acting opportunistically or in a passive way by
10
exiting the relationship or being neglectful. Partners satisfied with the relationship instead
appreciate the contacts with their counterparts, and the relationship likely is characterized by
trust, respect, and commitment (Ariño et al., 2001). In such a situation, managers should use
constructive response strategies (Brouthers and Bamossy, 2006), whether active or passive. For
example, Hibbard and colleagues (2001) argue that managers with positive views of a
relationship place less importance on an adverse situation and instead remain patient, believing
that the transient negative situation will improve. Geyskens and Steenkamp (2000) confirm that
socially satisfied managers are more likely to use creative or considerate voice.
Alliance-specific investments represent sunk costs that cannot be redeployed easily to another
alliance without some sacrifice in the productivity of the assets or cost to adapt them (Ping,
1993). These investments would be lost if the alliance were dissolved, so they act as exit barriers.
Their presence constitutes a source of dependence for the firm that makes them, which implies an
adverse situation for managers who need to reduce the negative consequences of their firms’
vulnerable position. The presence of unilateral, alliance-specific investments triggers constructive
response strategies and inhibits destructive ones (Hirschman, 1970; Tjemkes and Furrer, 2010),
because constructive responses reduce the risk of losing the investments if the relationship
terminates prematurely. In the case of high alliance-specific investments, constructive responses
may either be active, such as using creative or considerate voices to demonstrate commitment to
the relationship, or passive, such as being patient to reduce the negative effect of the dependence.
When alliance-specific investments are low though, managers have more latitude to act
destructively, because their lesser dependence on their partner means they may exit the
relationship, become more neglectful, or act opportunistically without fear of retaliation. The
results of previous response strategy research support this logic (e.g., Ping, 1993).
Finally, alternative availability refers to the extent to which the firm possesses attractive
11
alternatives outside the alliance that could enable it to attain its objectives (Ping, 1993). The
presence of attractive alternatives provides firms with a source of power, whereas a dearth of
alternatives increases dependence on counterparts. In an adverse situation without alternatives,
managers have strong incentives to make the current alliance work and likely respond actively to
improve the situation (Buchanan, 1992). For example, they might constructively use considerate
and creative voice to ensure their partner collaboration, but because they do not depend on their
partner, they also can use aggressive voice or opportunism if their partner is not cooperative.
Moreover, if managers perceive that they have other alternatives for achieving their objectives,
they depend less on the current relationship, which increases the likelihood of exit and passive
strategies such as neglect and patience (Ping, 1993; Tjemkes and Furrer, 2010).
Although in general empirical studies support the directions of the effects of the exchange
variables on response strategy preference, the strength of the relationship varies across studies.
By focusing solely on the effect of exchange variables, prior literature grounded in the investment
model implicitly has assumed that with a specific combination of exchanges variables, every
manager will prefer the same response strategies. For example, Tjemkes and Furrer (2010) in an
experimental study in the Netherlands find that economic satisfaction, social satisfaction,
alliance-specific investments, and the availability of attractive alternatives differentially and
interactively affect response strategies. However, in their study, they neglect contextual variables,
such as national culture, and their influence. In an international context, national culture likely
influences managers’ preference for a response strategy (e.g., Lee and Jablin, 1992; Morris et al.,
1998; Radford et al., 1993), as well as their perception of the adversity of the situation; thus, it
should moderate the effect of exchange variables (Thomas and Au, 2002; Thomas et al., 2003).
We therefore develop hypotheses related to both the direct and moderating effects of national
culture on response strategy preference.
12
3. Hypothesis development
To understand international differences in response strategy preference, we turn to Hofstede
(2001), who identifies four cultural dimensions: individualism–collectivism, masculinity–
femininity, power distance, and uncertainty avoidance. We focus on these cultural dimensions to
explain country differences (Taras, Rowney, and Steel, 2009) because they enable us to
disentangle theoretically the direct and interaction effects of a country on response strategy
preference (per Thomas and Pekerti, 2003). National culture and the country are not the same, yet
country offers a good proxy for national culture (Hofstede, 2001). We hypothesize specifically
that the individualism–collectivism and uncertainty avoidance cultural dimensions influence the
active–passive (not constructive–destructive) dimension of response strategy preference and that
the power distance and masculinity–femininity cultural dimensions influence the constructive–
destructive (not active–passive) dimension. In addition, we predict that the individualism–
collectivism and uncertainty avoidance cultural dimensions interact with two exchange variables,
economic satisfaction and alternative availability, in affecting the same active–passive dimension
of response strategies. Similarly, we hypothesize that the power distance and masculinity–
femininity cultural dimensions interact with social satisfaction and alliance-specific investments,
because they affect the same constructive–destructive dimension of response strategies.
Our hypotheses about the direct and moderating effects of national culture on response
strategy preferences concern fundamental cultural processes, which pertain to interpersonal
relationships. However, as we test them in an experimental ISA context, we frame them in such a
context. Moreover, as we use business students acting as alliance managers rather than actual
alliance managers to test these hypotheses, we refer to future managers when we present the
specific relationships we test.
13
3.1. Individualism–collectivism
The individualism–collectivism dimension refers to societal norms regarding individual–
group obligations and relationships. That is, individualism describes cultures in which the ties
between individuals are loose (Hofstede, 2001). Thus within an alliance context, responsibility
for action and responses to adverse situations lies within individual managers, who are more
likely to be active, using different forms of voice or acting opportunistically. In collectivist
cultures, the group is the dominant structure, and most actions reflect a consideration of their
effect on the group and its members (Hofstede, 2001), so responsibility for action and response
lies within the group. In collectivist cultures, decision-making processes and responses to
adversity are guided by consensus (Thomas et al., 2003), which results in more passive responses
(Yum, 2004). In such cultures, future managers likely prefer, more than those in individualistic
countries, to exit the relationship or be neglectful or patient (Pressey and Qu, 2007).
In countries with an individualistic culture, decision-making processes and responses to
adverse situations reflect motives pertaining to the protection of individual profits, as justified by
utilitarian principles (Thomas et al., 2003). Thus, if the relationship is perceived as valuable,
future managers in individualist countries will prefer constructive strategies. However, when the
relationship loses its value, they will act destructively. Compared with individualistic countries,
people in countries with a collectivist culture are inclined to consider their business exchange
partners as out-groups (Nakana, 1971), especially if they are foreigners (Johnson et al., 1996).
Therefore, in an adverse situation, they will not hesitate to use destructive strategies to protect
group harmony and save face. However, when responding constructively is in the best interest of
the group, they will do so to maintain the status quo and limit actions that might disrupt in-group
harmony (Triandis, 1995). Thus, in an alliance context, future managers in both individualistic
and collectivistic countries should prefer constructive and destructive responses in similar ways.
14
However, because the individualism–collectivism dimension affects active–passive and not
constructive–destructive response preferences, we hypothesize that within the context of alliances
in individualist countries, active response strategies are relatively more preferred than passive
response strategies, compared with collectivistic countries.
Hypothesis 1a: In individualist countries, active response strategies (opportunism, aggressive,
creative, and considerate voices) are preferred relatively more and passive strategies
(patience, neglect, and exit) relatively less than in collectivistic countries.
Also in relation to individualism–collectivism, economic satisfaction influences the degree of
activeness and passiveness in response preference, such that these factors likely interact.
Achieving economic satisfaction is a more important goal in individualistic countries than in
collectivist countries, because alliances in the former are governed by more rational cost–benefit
calculations (Triandis, 1995). Therefore, when economic satisfaction decreases, future managers
in individualistic countries are more likely than their counterparts in collectivist countries to
prefer active response strategies and less likely to prefer passive strategies to solve the situation
(Thomas and Au, 2002). In contrast, future managers in collectivist countries are less sensitive to
changes in economic satisfaction, as the quality of the relationship with their partner is more
important than its short-term financial outcomes. Therefore:
Hypothesis 1b: In individualistic countries, the negative effect of economic satisfaction on active
response strategies and the positive effect of economic satisfaction on passive strategies are
weaker than they are in collectivistic countries.
Similar to the effect of economic satisfaction, the presence of attractive alternatives should
influence the degree of activeness and passiveness of future managers’ response strategy
preference, so we expect alternative attractiveness to interact with the individualism–collectivism
cultural dimension. In individualistic countries, which better tolerate self-serving behaviors and
15
relationships based on cost–benefit calculations (Triandis, 1995), future managers without
attractive alternatives feel threatened by the risk that their counterpart will prefer to exit the
alliance; to reduce their dependence (Thomas and Au, 2002), they prefer active strategies more
and passive strategies less than managers in collectivistic countries. Future managers from
countries with collectivist cultures instead are used to depending on their group and therefore
might feel less threatened by a dependence situation created by a lack of alternatives. Therefore,
they are likely to be less influenced by the existence or absence of alternatives, because they do
not consider the situation especially adverse. Thus, we hypothesize:
Hypothesis 1c: In individualistic countries, the negative effect of alternative availability on
active response strategies and the positive effect of alternative availability on passive
strategies are stronger than they are in collectivistic countries.
3.2. Masculinity–femininity
The masculinity–femininity cultural dimension pertains to norms regarding an achievement
motivation versus quality of life. Masculine cultures convey norms that emphasize the need for
autonomous, competitive, and assertive actions to achieve materialistic goals (Hofstede, 2001;
Hofstede and Usunier, 2003). In contrast, in countries with a feminine culture, the dominant
norms emphasize collaboration, relationships, and caring for others (Hofstede, 2001; Hofstede
and Usunier, 2003). Masculine cultural norms favor ego-boosting behaviors and sympathy for the
strong, which suggests a preference for destructive, less constructive behaviors. In such countries,
conflicts get resolved through fighting, in an effort to “get things straight.” Thus, in an alliance
context, future managers from masculine countries likely prefer more destructive response
strategies compared with future managers from feminine countries. In contrast, feminine cultural
norms exhibit a pattern of nurturance, which emphasizes less destructive and more constructive
16
behavior (Doney et al., 1998). Feminine cultural norms also emphasize modesty (Hofstede,
2001), as reflected in a preference for compromises and less destructive response strategies, such
as creative and considerate voices and patience (Doney et al., 1998; Hofstede and Usunier, 2003).
However, the masculine–feminine cultural dimension does not influence the degree of
activeness–passiveness of future managers’ response strategy preference. In masculine countries,
destructive strategies are preferred over constructive ones, independent of their degree of
activeness or passiveness. These countries value competition and assertive actions (Hofstede,
2001; Hofstede and Usunier, 2003), so future managers are more likely to use aggressive voice,
act opportunistically, exit the relationship, or act neglectfully than use constructive strategies,
compared with those in feminine countries. In contrast, in feminine countries, in which the
dominant norms emphasize collaboration and relationships (Hofstede, 2001; Hofstede and
Usunier, 2003), future managers use constructive strategies, such as creative and considerate
voices and patience, independent of the degree of activeness–passiveness, more so than
destructive strategies, in comparison with those in masculine countries. Thus, we propose that the
masculine–feminine cultural dimension affects the constructiveness–destructiveness of future
managers’ response strategy preferences, not the activeness–passiveness of this preference.
Hypothesis 2a: In masculine countries, destructive response strategies (neglect, exit,
opportunism, and aggressive voice) are preferred relatively more and constructive strategies
(creative and considerate voices, and patience) relatively less than in feminine countries.
Because social satisfaction and the masculinity–femininity cultural dimension both influence
the degree of constructiveness versus destructiveness, we posit that they interact. In masculine
countries, social harmony and cooperative relationships are not goals in themselves, and fighting
is a means to clarify a situation. Thus, future managers likely prefer destructive strategies over
constructive strategies, compared with future managers in feminine cultures, irrespective of their
17
level of social satisfaction. In contrast, social harmony and good relationship quality are critically
important in feminine countries, which tolerate socially dissatisfying relationships far less than
do masculine cultures (Hofstede and Usunier, 2003). Therefore in feminine countries, social
satisfaction should have a stronger effect on response strategy preference than it does in
masculine cultures. As social satisfaction increases, future managers from feminine countries,
who value consensus and close relationships (Hofstede, 2001), respond more constructively,
whereas when social satisfaction decreases, they likely respond more destructively. Thus:
Hypothesis 2b: In masculine countries, the positive effect of social satisfaction on constructive
response strategies and the negative effect of social satisfaction on destructive strategies are
weaker than they are in feminine countries.
Similar to the effect of social satisfaction, the presence of alliance-specific investments
should influence the degree of constructiveness–destructiveness of future managers’ response
strategy preference, so we expect alliance-specific investments to interact with the masculinity–
femininity cultural dimension. As mentioned, feminine countries value close relationships
(Hofstede, 2001), so future managers who have made alliance-specific investments do not
perceive their dependence on their partner as a threat and are more likely than those from
masculine countries to prefer constructive responses that protect the quality of their relationship,
even if such behavior could increase their dependence. Therefore in feminine countries, future
managers are more likely to prefer constructive responses and less likely to prefer destructive
responses than in masculine cultures, irrespective of the amount of their alliance-specific
investments. In contrast, when future managers from masculine countries perceive alliancespecific investments as high, they likely feel threatened by their dependence on their counterpart,
and to escape this dependence, they tend to prefer destructive response strategies, not
constructive responses, relative to situations with low alliance-specific investments. Thus, we
18
hypothesize:
Hypothesis 2c: In masculine countries, the positive effect of alliance-specific investments on
constructive response strategies and the negative effect of alliance-specific investments on
destructive strategies are weaker than they are in feminine countries.
3.3. Power distance
Power distance pertains to norms for the acceptable distribution of power in a society, that is,
the degree to which less powerful members within a society expect and accept that power is
distributed unequally (Hofstede, 2001). Inequality in small power distance countries is a
necessary evil that should be minimized; in large power distance countries, inequality is the basis
of societal order (Hofstede, 2001). Therefore, in large power distance countries, the use of power
through destructive strategies, such as exiting the relationship, opportunism, aggressive voice, or
neglect, demands less legitimization than it would in small power distance countries. In small
power distance countries though, the dominant cultural norm of harmony between the powerful
and powerless favors constructive response strategies, such as creative and considerate voices and
patience, because they are more likely to preserve the relationship between alliance partners.
Because power distance relates to the distribution of power within a society, which can take
both active and passive (i.e., potential power) forms, the power distance cultural dimension is not
likely to influence the activeness or passiveness of future managers’ response strategy preference.
Specifically, in large power distance countries, future managers may prefer both active and
passive destructive response strategies over constructive ones, often by legitimizing neglect and
aggressive voice. In small power distance countries, patience and considerate and constructive
voices are all constructive strategies, whether active or passive, that can protect harmony. Thus,
the power distance cultural dimension likely affects the constructiveness–destructiveness of
19
response strategy preference but not the active–passive dimension:
Hypothesis 3a: In high power distance countries, destructive response strategies (neglect, exit,
opportunism, and aggressive voice) are preferred relatively more and constructive strategies
(creative and considerate voices, and patience) relatively less than in low power distance
countries.
Because both social satisfaction and the power distance cultural dimension influence the
degree of constructiveness versus destructiveness of the response strategy preference, we posit
that they interact. As in masculine countries, in large power distance countries, social harmony
and cooperative relationships are not goals in themselves. Future managers in such countries
therefore are more likely to prefer destructive strategies and less likely to prefer constructive
strategies compared with future managers in small power distance countries, regardless of their
level of social satisfaction. In contrast, social harmony and good relationship quality are critically
important in small power distance countries, which tolerate socially dissatisfying relationships far
less than do large power distance countries (Hofstede, 2001). Therefore in small power distance
countries, social satisfaction should have a stronger effect on response strategy preference than in
large power distance cultures. As social satisfaction increases, future managers from small power
distance countries likely respond more constructively, whereas when social satisfaction
decreases, they are likely to respond more destructively. Thus:
Hypothesis 3b: In high power distance countries, the positive effect of social satisfaction on
constructive response strategies and the negative effect of social satisfaction on destructive
strategies are weaker than in low power distance countries.
Similar to the effect of social satisfaction, the presence of alliance-specific investments can
influence the degree of constructiveness and destructiveness of future managers’ response
strategy preference, so we expect alliance-specific investments to interact with the power distance
20
cultural dimension. Power distance deals with the need for dependence versus interdependence in
society (Hofstede, 2001). In high power distance countries, people strive for independence; in
small power distance countries, they likely accept interdependence with partners. Therefore,
when the amount of alliance-specific investments increases, future managers in large power
distance countries are likely to react more constructively and less destructively to reduce the risks
related to their dependence. On the contrary, in countries with small power distance, when the
amount of alliance-specific investments increases, future managers do not feel more threatened,
nor do they respond differently than they would in situations with low levels of alliance-specific
investments.
Hypothesis 3c: In high power distance countries, the positive effect of alliance-specific
investments on constructive response strategies and the negative effect of alliance-specific
investments on destructive strategies are weaker than in low power distance countries.
3.4. Uncertainty avoidance
Uncertainty avoidance pertains to the degree to which the members of a society feel
uncomfortable with uncertainty and ambiguity (Hofstede, 2001). In an alliance context, future
managers in high uncertainty countries seek to reduce uncertainty by engaging in passive
response strategies, rather than active ones, to reduce the risk of harmful partner retaliatory
behavior. Passive response strategies, such as patience and neglect, or even exit, are perceived as
less uncertain than more active ones, such as different types of voice and opportunism. In
contrast, we expect that future managers from higher uncertainty countries prefer more active
response strategies, which may result in more uncertain outcomes. In countries with low
uncertainty avoidance, people are more willing to accept uncertainty and not afraid to take risks.
21
Uncertainty avoidance is less likely to influence the degree of constructiveness–
destructiveness of future managers’ response strategy preference, because active strategies seem
more uncertain, irrespective of their constructiveness or destructiveness. For example,
opportunism may trigger retaliatory behavior, but creative voice is also ambiguous, because it
challenges the status quo. In contrast, constructive and destructive passive strategies appear less
risky, because they allow for more time and their effects emerge only in the longer term.
Therefore, the degree of constructiveness–destructiveness is not likely to be influenced by the
uncertainty avoidance cultural dimension. We postulate that uncertainty avoidance instead affects
the activeness–passiveness of future managers’ response strategy preferences.
Hypothesis 4a: In countries with high uncertainty avoidance, active response strategies
(opportunism, aggressive, creative, and considerate voices) are preferred relatively less and
passive strategies (patience, neglect, and exit) relatively more than in countries with low
uncertainty avoidance.
Economic satisfaction also influences the degree of activeness and passiveness in response
preference, so these two factors should interact. Achieving short-term economic performance is a
more important goal in high uncertainty avoidance countries than in low uncertainty avoidance
countries. Therefore, when economic satisfaction is low, future managers in high uncertainty
avoidance cultures are more likely than those in low uncertainty avoidance countries to prefer
active and less likely to prefer passive response strategies to reduce the ambiguity created by low
economic performance. In contrast, future managers in low uncertainty avoidance countries are
less sensitive to changes in economic satisfaction, because they are less afraid of uncertainties
and ambiguities, such that they may trade off short-term performance losses for long-term
benefits.
Hypothesis 4b: In high uncertainty avoidance countries, the negative effect of economic
22
satisfaction on active response strategies and the positive effect of economic satisfaction on
passive strategies are stronger than in low uncertainty avoidance countries.
Finally, and similar to the effect of economic satisfaction, the presence of attractive
alternatives should influence the degree of activeness and passiveness of future managers’
response strategy preference and interact with the uncertainty avoidance cultural dimension. In
high uncertainty avoidance countries, future managers without attractive alternatives feel
threatened that their counterpart will prefer to exit the alliance, so to reduce the ambiguity and
uncertainty related to their dependence, they prefer active strategies more and passive strategies
less, compared with future managers from low uncertainty avoidance countries. In contrast, those
from low uncertainty countries are less threatened by the uncertainties of a dependence situation
created by a lack of alternatives. They are then less influenced by the availability of attractive
alternatives, and we hypothesize:
Hypothesis 4c: In high uncertainty avoidance countries, the negative effect of alternative
availability on active response strategies and the positive effect of alternative availability on
passive strategies are stronger than in low uncertainty avoidance countries.
4. Methodology
To test the hypotheses empirically, we designed a scenario-based experiment, a method that
has proven useful for studying response strategies (Lee and Jablin, 1992; Rusbult et al., 1988;
Tjemkes and Furrer, 2010) and international management decisions (Joardar et al., 2007).
Furthermore, with an experiment, we can examine causal relationships between predictors and
outcomes in controlled conditions (i.e., reduce impact of confounding effects), which makes this
method particularly suitable for disentangling direct and moderating effects (Bateman and
Zeithaml 1989; Croson et al., 2007).
23
4.1. Sample
To assess the effects of national culture differences on response strategies, we collected data
in five countries: Japan, the Netherlands, Switzerland (French-speaking part), Turkey, and the
United Kingdom. These five countries systematically vary according to the four cultural
dimensions of Hofstede, ranging from 37 to 89 for individualism, from 14 to 95 form
masculinity, from 35 to 70 for power distance and from 35 to 92 for uncertainty avoidance (see
Table 1). We focus on Hofstede’s dimensions rather than others, given evidence that it has had
far greater impact (Kirkman et al, 2006). In addition, even if Hofstede’s scores have been
criticized for being outdated and collected in a specific context, recent meta-analysis (Taras et al.,
in press) and data (Ralston et al., in press) show that if country scores tend to converge over time,
the relative position of our five countries is relatively stable.
[Insert Table 1 about here]
We collected 1,379 questionnaires from business students enrolled in masters’ programs
(M.Sc.) in Japan, the Netherlands, Switzerland, Turkey, and the United Kingdom. Only students
born and raised in their country were retained in the final samples. The Dutch sample consisted of
334 students, with an average age of 24.4 years, 35.5% of whom were women. The Swiss sample
consisted of 255 business students with an average age of 23.3 years, 42.4% of whom were
women. The Turkish sample consisted of 278 students, with an average age of 23.3 years, 49.6%
of whom were women. The Japanese sample consisted of 262 respondents with an average age of
24.4 years, 32.1% of whom were women. The U.K. sample consisted of 250 students with an
average age of 23.7 years, and 46% of them were woman.
As our hypotheses are not specific to ISAs but to more general interpersonal relationships
and pertain to relative stable characteristics of human nature (Bello et al., 2009), two arguments
support our decision to use business students, as respondents. First, the use of a homogeneous
24
sample of business students allows us to improve internal validity (Croson et al., 2007). Second,
compared to actual managers, whose responses may be contaminated by past experience with a
specific existing alliance (framing bias), business students are more likely to only respond to the
manipulations (Bateman and Zeithaml 1989). In addition, as they represent future managers, they
are likely to understand the strategic alliance context of the experiment.
As already mentioned the use of business students may raise the issue of external validity,
therefore to assess the validity of our results, we collected additional survey data from a sample
of 135 Dutch alliance managers (average age = 42.6 years [SD = 9.0]; average alliance
management experience = 4.1 [SD = 1.8] on seven-point experience scale; 69.6% men), and
compared their results with those of the Dutch business students. This comparison allows us to
test the validity of our measurement and manipulations.
The questionnaire, originally developed in English, was translated into Japanese and Turkish,
following standard translation and back-translation procedures. Students in the Netherlands,
Switzerland, and the United Kingdom used the English version of the questionnaire, because
English was the language used in their study programs.
4.2. Scenario-based experiment
To capture adverse situations, we manipulated economic satisfaction, social satisfaction,
alliance-specific investments, and the availability of alternatives (per Tjemkes and Furrer, 2010).
The experiment in each country used a four-factor (exchange variables) by two level (positive
versus negative) between-subjects design, in which we combined the manipulations to form 16
different scenarios, though we removed the all-positive (i.e., not adverse) scenario. Each
experiment was conducted over multiple sessions, in which we assigned each participant
randomly to one of the 15 scenarios. The participants assumed the roles of alliance managers and
25
indicated how they would respond to the adverse situation presented in the scenario. The text of
the scenarios appears in the Appendix.
To manipulate economic satisfaction, we varied the financial outcomes of the alliance, such
that it produced either beyond or below their expectations. The social satisfaction manipulation
centered on relationship quality. The quality of the relationship referred to the extent to which the
firms developed a relationship characterized by trust, commitment, and adaptability, which
eliminated the need for substantial monitoring costs. For the alliance-specific investment
manipulation, we varied the extent to which the focal firm made substantial investments in the
alliance and could expect switching costs. Specifically, the focal firm either invested little or
substantially and could or could not expect additional costs from penalty fees if the alliance
ended prematurely. Finally, we manipulated the availability of alternatives as the extent to which
the focal firm had other means to realize its objectives, whether many or only a few available
alliance partners.
4.3. Response strategy measures
To operationalize the seven response strategies, we turned to existing scales and, when
necessary, adapted them to the context of ISA. We measured exit with items pertaining to
whether the respondent intended to end the relationship or stop doing business with the partner
(Rusbult et al., 1982; Withey and Cooper, 1989). The measures for opportunism, adapted from
Ping (1993) and Wathne and Heide (2000), included withholding information, exaggerating the
adverse nature of the situation, and escaping from contractual obligations. Aggressive voice items
referred to pushing a solution forcefully or being persistent (Hibbard et al., 2001). For creative
voice, we used items related to the creation of innovative and creative solutions or fresh ideas
(Zhou and George, 2001). To measure considerate voice, the items indicated working to create a
26
consensus and finding a solution satisfactory and acceptable for everyone (Ping, 1993). We
operationalized patience with items such as optimistically waiting for better times and trusting
that the situation would resolve itself (Hibbard et al., 2001; Ping, 1993). For neglect, the items
referred to not dealing with the issue, not putting additional effort into the relationship, and not
presenting initiatives to improve the situation (Ping, 1993). All these measures used seven-point
Likert scales, ranging from “I would definitely not react in this way” [1] to “I would definitely
react in this way” [7]. The list of items appears in Table 2.
[Insert Table 2 about here]
4.4. Control variables
To take into account the potential effect of noncultural factors, we also measured several
control variables: A single-item scale assessed perceptions of the severity of the situation, which
can influence the choice of response strategies (Rusbult et al., 1988). In addition, because some
response strategies may be more socially desirable (considerate voice) than others (opportunism),
we included the M-C2 version of the Marlowe-Crowne Social Desirability scale (Strahan and
Gerbasi, 1972). Finally, we measured the respondents’ age and gender, because these personal
characteristics may influence preferences for response strategies (e.g., Rusbult et al., 1988).
5. Analyses and results
We conducted four sets of analyses to examine the data. First, we assessed the reliability and
construct validity of the response strategy measures using Cronbach’s alpha and the composite
reliability (CR) coefficients, as well as the factor loadings from confirmatory factor analyses
(CFAs) conducted in each country separately. Second, we ran a series of multigroup CFAs to
assess the degree to which the seven-factor model is invariant across the four countries. Third, to
examine the direct and moderating effects of country differences and test the hypotheses, we
27
conducted a multivariate analysis of covariance (MANCOVA), followed by a series of post-hoc
country comparisons and F-tests. Because the response strategies were interrelated, we
manipulated the scenario variables, and we used covariates to control for confounding effects, a
MANCOVA was the most appropriate method (Huberty and Morris, 1989). Fourth, to examine
the external validity of the findings, we collected additional data from alliance managers in the
Netherlands and assessed the measurement equivalence and similarity of their responses with
those from the business students in our Dutch sample.
5.1. Construct validity and cross-cultural invariance of the response strategies
We conducted a separate CFA for each country to test the proposed seven-factor response
strategy structure. The alphas and CR greater than or equal to .70, along with factor loadings that
exceed .50, suggest acceptable reliability and convergent validity (Nunnally and Bernstein,
1994). We tested seven-factor models rather than the seven dimensions separately so that we
could investigate the discriminant validity of the seven response strategies simultaneously
(Perrinjaquet et al., 2007). We employed maximum likelihood estimation procedures, because the
data do not strongly violate multivariate normality assumptions (McDonald and Ho, 2002).
Following common practice (e.g., Byrne, 2001; Hu and Bentler, 1999), we also used multiple
indicators to assess model fit, namely, the normed chi-square (χ2/d.f.), root mean square error of
approximation (RMSEA), standardized root mean square residual (SRMR), non-normed fit index
(NNFI), and comparative fit index (CFI), and we required RMSEA ≤ .06, SRMR ≤ .08, NNFI ≥
.90, and CFI ≥ .95, as well as χ2/d.f. less than or equal to 2, to confirm good model fit.
We used AMOS 17.0 to estimate a seven-factor CFA model with the 21 response strategy
items for each country. The analyses reveal no offending estimates in any of the models. The
country models possess good fit (see Table 3); the normed chi-square values are 1.53, 1.50, 1.54,
28
1.98, and 1.57, for Japan, the Netherlands, Switzerland, Turkey, and the United Kingdom
respectively. In addition, other goodness-of-fit indices suggest acceptable fit: The RMSEA values
range from .031 [90% confidence interval (CI): .013, .044] for the United Kingdom to .059 [90%
CI: .050, .069] for Turkey, below the cut-off value. The other indices also suggest a good fit; for
Japan, the fit indices are .058 (SRMR), .95 (NNFI), and .96 (CFI); for the Netherlands, these
values are .049 (SRMR), .96 (NNFI), and .97 (CFI); for Switzerland, they are .052 (SRMR), .95
(NNFI), and .96 (CFI); for Turkey, .059 (SRMR), .90 (NNFI), and .92 (CFI); and for the United
Kingdom, they are .046 (SRMR), .95 (NNFI), and .96 (CFI). The Turkish CFI is thus slightly
below the expected value but sufficient.
[Insert Table 3 about here]
To assess convergent validity, we examined the factor loadings, which are significant and
exceed the .50 threshold, ranging from .53 to .85 in the Japanese sample, .56 to .90 in the Dutch
sample, .51 to .89 in the Swiss sample, .55 to .88 in the Turkish sample, and .56 to .90 in the
British sample, with one exception. The Cronbach’s alphas and CR values are greater than .70,
with a few exceptions that are still above .60 (see Table 4). The average variances extracted are
slightly below their expected values (.34–.73), but the square roots range from .58 to .85, higher
than any of their respective pairwise correlations, indicating discriminant validity.
[Insert Table 4 about here]
We then tested for measurement and construct invariance to determine whether the response
strategies remain invariant across countries. Measurement invariance pertains to the psychometric
properties of the measurement scales and includes configural invariance, metric invariance, and
scalar invariance. Scalar invariance is a prerequisite for interpreting construct differences
(Steenkamp and Baumgartner, 1998). However, full scalar invariance is not necessary, provided
at least one item per response strategy is invariant (i.e., partial invariance) (Byrne, 2001). To
29
evaluate measurement and construct invariance, we used multigroup CFA models and considered
group comparisons across the five countries. Full scalar invariance is not achieved, but after
releasing four factor loadings and nine intercept constraints, we establish partial scalar invariance
with satisfactory fit indices: The normed chi-square value is 1.53, RMSEA is .020 [90% CI: .018,
.022], SRMR is .049, NNFI is .94, and CFI is .95. The comparative fit indices between an
unconstrained model and the partial invariant model are not significantly different (∆χ2 = 44.5,
p = .11, ∆CFI = .001), in support of partial scalar invariance. These results enable us to conduct
mean comparisons across countries to test the hypotheses.
5.2. Impact of cultural differences on response strategy preference
To test the effect of country on response strategies, we conducted a MANCOVA. We ran
post-hoc group comparisons and F-tests to test the hypotheses and interpret the effects of country
differences on response strategies, as well as the impact of the exchange variables. For our
analyses, we used the average scores for each response strategy as dependent variables and the
scenario manipulations and country as the fixed factors, with gender, age, problem severity, and
social desirability as covariates. Prior to the analysis, we examined the MANCOVA assumptions
and found no violations.
The omnibus MANCOVA tests indicate significant differences for the exchange variables
and countries. Specifically, we find significant Wilks’ lambdas for country (Λ = .77, F = 12.96, p
< .001) and the four exchange variables: economic satisfaction (Λ = .96, F = 7.93, p < .001),
social satisfaction (Λ = .93, F = 13.76, p < .001), alliance-specific investments (Λ = .99, F = 2.65,
p < .01), and alternative availability (Λ = .92, F = 16.80, p < .001). Of the four hypothesized twoway interactions between country and exchange variables, three are significant: with economic
satisfaction (Λ = .96, F = 2.20, p < .001), social satisfaction (Λ = .96, F = 1.81, p < .01), and
30
alternative availability (Λ = .96, F = 2.06, p < .01). The interaction between country and alliancespecific investment is not significant. Furthermore, three of the four control variables are
significant: age (Λ = .99, F = 2.13, p < .05), problem severity (Λ = .98, F = 3.17, p < .01), and
social desirability (Λ = .97, F = 5.69, p < .05). Gender is not significant (Λ = .99, F = .70, p =
.67). The F-values of the corrected model, which reflect variations in the response strategies
attributable to country, exchange variables, and covariates, show significant results for all seven
response strategies (see Table 5).
[Insert Table 5 about here]
To test the hypotheses, we conducted post-hoc F-tests. With regard to the direct effect of
country, the univariate F-tests show significant differences for all seven strategies (see Table 5):
exit (F = 22.53, p < .001), opportunism (F = 20.27, p < .001), aggressive voice (F = 15.79, p <
.001), creative voice (F = 11.96, p < .01), considerate voice (F = 8.74, p < .001), patience (F =
11.09, p < .001), and neglect (F = 9.04, p < .001).
Hypothesis 1a proposes that in individualist countries, active response strategies are
preferred relatively more and passive strategies relatively less than in collectivistic countries. The
results of the pairwise comparisons in Table 5 suggest that the hypothesis is fully supported for
creative and considerate voices, patience, and neglect. Overall, respondents in the United
Kingdom, the Netherlands, and Switzerland—the more individualist countries—are more likely
to use creative and considerate voices and less likely to use patience and neglect than respondents
in Japan and Turkey, the more collectivist countries. Hypothesis 1a is also partially supported for
opportunism (except for Japan) and aggressive voice (except for Switzerland), but it is not
supported for exit. Overall then, we find support for Hypothesis 1a.
Hypothesis 2a proposes that in masculine countries, destructive response strategies are
31
preferred relatively more and constructive strategies relatively less than in feminine countries.
The results of the pairwise comparisons suggest support for the hypothesis for exit and creative
and considerate voice in the case of Japan, the most masculine country in our sample, and for
considerate voice and neglect in the case of the Netherlands, the most feminine country. For the
United Kingdom, the second most masculine country in our sample, we find some support for the
hypothesis in relation to exit, opportunism, aggressive voice, patience, and neglect. The results
are also consistent with our expectations for Switzerland, which scores in the middle in terms of
masculinity. However, the results for Turkey conflict with our expectations. Overall, Hypothesis
2a receives support.
Hypothesis 3a proposes that in large power distance countries, destructive response strategies
are preferred relatively more and constructive strategies relatively less than in small power
distance countries. The results of the pairwise comparisons confirm the hypothesis for
opportunism, considerate voice, and neglect in the case of Turkey, a country with a large power
distance. The hypothesis is also supported for creative and considerate voice in the case of the
Netherlands and the United Kingdom, the two countries with the smallest power distance in our
sample. For creative and considerate voices, the results for Switzerland are contrary to our
expectations, in that respondents in this country scored relatively high. Overall, we uncover
partial support for Hypothesis 3a.
Hypothesis 4a proposes that in countries with high uncertainty avoidance, active response
strategies are preferred relatively less and passive strategies relatively more than in countries with
low uncertainty avoidance. The results of the pairwise comparisons suggest that the hypothesis is
fully supported for aggressive, creative, and considerate voices, patience, and neglect. Overall,
respondents in the Netherlands and the United Kingdom, the countries with the smallest
uncertainty avoidance, are more likely to use aggressive, creative, and considerate voices and less
32
likely to use patience and neglect than are respondents in Japan, Turkey, and Switzerland, the
countries with the largest uncertainty avoidance. Hypothesis 4a is also partially supported for
opportunism (except for Japan and Turkey). Overall, we find support for Hypothesis 4a.
Consistent with our expectations and across countries, the exchange variables influence
response strategy preference. Economic satisfaction has a direct negative effect on exit (F =
36.82, p < .001) and a positive effect on patience (F = 10.53, p < .001). Social satisfaction has
direct negative effects on exit (F = 67.20, p < .001), opportunism (F = 5.71, p < .05), and
aggressive voice (F = 8.07, p < .01) and positive effects on considerate voice (F = 10.01, p < .01)
and patience (F = 11.93, p < .001). Alliance-specific investments have direct negative effects on
exit (F = 5.84, p < .05) and aggressive voice (F = 3.83, p < .05) and a positive effect on
considerate voice (F = 11.78, p < .001). The lack of attractive alternatives, across countries, has a
direct negative effect on exit (F = 92.31, p < .001) and neglect (F = 8.77, p < .01) and a positive
effect on creative voice (F = 32.11, p < .001), considerate voice (F = 60.77, p < .001), and
patience (F = 8.77, p < .01).
The interaction effects between the country and economic satisfaction are statistically
significant for considerate voice (F = 4.03, p < .01) and patience (F = 3.88, p < .01) but not for
the other response strategies. We provide the interaction plot for considerate voice in Figure 1a
and that for patience on Figure 1b. For these two response strategies, the results provide support
for Hypothesis 1b, except for Switzerland. In the Netherlands and the United Kingdom, two
individualistic countries, the positive effects of economic satisfaction on considerate voice and its
negative effects on patience are stronger than they are for Turkey and Japan, the two most
collectivist countries in our sample. The results also support Hypothesis 4b; for Japan and
Turkey, two countries with high uncertainty avoidance, the positive effects of economic
33
satisfaction on considerate voice and its negative effects satisfaction on patience are stronger than
for the low uncertainty avoidance countries, the Netherlands and the United Kingdom.
The interaction effects between country and social satisfaction are statistically significant for
exit (F = 3.62, p < .01) and opportunism (F = 2.47, p < .01) and not significant for the other
response strategies. The interaction plot for exit appears in Figure 1c; that for opportunism is in
Figure 1d. The results provide partial support for Hypothesis 2b, because the more feminine
Netherlands indicates a stronger negative effect of social satisfaction on exit than Turkey, with its
medium level of masculinity, and the United Kingdom, with its very masculine culture. However,
the significant effect of social satisfaction in Japan, a masculine culture, on exit contrasts with our
expectations. In addition, for opportunism, social satisfaction has the strongest effect in the
Netherlands compared with the more masculine countries, in line with our expectations.
However, contrary to our expectations, the effects of social satisfaction in Turkey and
Switzerland are weaker than they are in Japan and the United Kingdom. The results also provide
partial support for Hypothesis 3b: For Japan, Turkey, and Switzerland, the three countries with
the largest power distances in our sample, the negative effect of social satisfaction on exit is
stronger than it is for the United Kingdom, the country with smallest power distance, but it is not
stronger than that for the Netherlands.
The interaction effects between country and alternative availability are statistically
significant for exit (F = 4.71, p < .001), considerate voice (F = 2.46, p < .05), patience (F = 2.91,
p < .05), and neglect (F = 2.40, p < .05) but not for the other response strategies. We provide the
interaction plot for exit in Figure 1e and that for considerate voice in Figure 1f, with the plots for
patience in Figure 1g and for neglect in Figure 1h. For these four response strategies, the results
support Hypothesis 1c. In the Netherlands and Switzerland, two individualistic countries, the
negative effect of alternative availability on considerate voice and its positive effect on exit are
34
stronger than those for Turkey, the most collectivist country in our sample. In addition, the
positive effects of alternative availability on patience and neglect are stronger for the United
Kingdom, the most individualist country in our sample, than for Turkey. The results also provide
support for Hypothesis 4c by showing that for Japan, the country with the highest uncertainty
avoidance, the negative effect of alternative availability on considerate voice and its positive
effects on exit, patience, and neglect are stronger than the United Kingdom, the country with
lowest uncertainty avoidance.
[Insert Figure 1 about here]
Among the control variables, age, problem severity, and social desirability have significant
effects on response strategy preference. Age has significant effects on creative voice (F = 4.07, p
< .05) and considerate voice (F = 5.67, p < .05); the post-hoc tests reveal that older respondents
are more likely to use creative and considerate voices than are younger ones. Problem severity
influences exit (F = 3.87, p < .05) and patience (F = 10.80, p < .001); the more severe the
perception of the situation, the more likely respondents are to exit and the less likely they are to
be patient. Social desirability has a significant effect on preferences for several response
strategies: opportunism (F = 24.87, p < .001), creative voice (F = 8.06, p < .01), considerate
voice (F = 16.33, p < .001), and neglect (F = 3.85, p < .05). Respondents with high scores on the
social desirability scale are less likely to act opportunistically and neglectfully and more likely to
choose creative and considerate voices than are respondents with low social desirability scores.
The effect of gender is not significant though.
5.3. External validity test
To assess the external validity of our measurement scales and scenario manipulations, we
compared the results of the Dutch managers and business students in two steps. First, we used
35
multigroup CFA to test measurement invariance across the two groups. The results exhibit full
metric and partial scalar invariance (four intercepts released) with satisfactory fit indices: The
normed chi-square value is 1.54, RMSEA is .034 [90% CI: .028, .039], SRMR is .077, NNFI is
.94, and CFI is .95. The comparative fit indices between the unconstrained model and the partial
scalar equivalent model are not statistically significant (∆χ2 = 17.0, p = .11, ∆CFI = .002), in
support of partial scalar measurement invariance between the future and actual manager samples.
Second, we ran a MANCOVA with the seven response strategies as dependent variables; the
four exchange variables as fixed factors; a fixed factor sample to denote future and actual
managers; and age, gender, problem severity, and social desirability as covariates. We also
included two-way interactions in the model. If the sample and sample–exchange variable
interactions are not significant, we can confirm the external validity of the business student
sample, because no difference in response strategy preferences would exist between the future
and actual manager samples. Consistent with our expectations, the omnibus MANCOVA tests
indicate significant differences for the four exchange variables, but not for the sample factor.
Specifically, we find significant Wilks’ lambdas for economic satisfaction (Λ = .95, F = 3.27, p <
.01), social satisfaction (Λ = .89, F = 7.71, p < .001), alliance-specific investments (Λ = .97, F =
2.29, p < .05), and alternative availability (Λ = .96, F = 2.80, p < .01) but not for the sample
factor (Λ = .99, F = .70, p = .68). The four interactions between the sample factor and the
exchange variables are not significant at 5%; the only significant two-way interaction is between
the sample factor and problem severity (Λ = .97, F = 2.29, p < .05) for two response strategies:
patience (F = 5.84, p < .05) and neglect (F = 11.35, p < .001). That is, actual managers are more
patient and neglectful than future managers when problem severity increases. Overall, these
results provide support for the external validity of our measures and experimental design.
36
6. Discussion
This study investigates cross-country differences in preferences for using seven response
strategies in international strategic alliances. A scenario-based experiment among 1,379 students
acting as alliance managers from five countries—Japan, the Netherlands, Switzerland, Turkey,
and the United Kingdom—demonstrates that country directly influences the preference for
response strategies and moderates the effects of economic and social satisfaction and the
availability of alternatives.
6.1. Interpretation of main results
Respondents in Japan, a country characterized by high scores on masculinity and uncertainty
avoidance and medium scores on individualism and power distance, are more likely to use exit
and patience; less likely to use opportunism, creative voice, and considerate voice; and indifferent
toward aggressive voice and neglect compared with respondents from other countries. This
summary indicates that for Japanese respondents, preferences for destructive response strategies
and the rejection of constructive response strategies likely reflects a high level of masculinity
and, to some extent, a medium level of power distance. Although the results for the passive
responses cannot be explained by the direct effects of the cultural dimensions, the country
interactions with economic satisfaction and alternative availability suggest that sensitivity to
these exchange variables, due to high uncertainty avoidance, affects response strategy
preferences. Similarly, high masculinity makes Japanese respondents less sensitive to social
satisfaction, which explains their low preference for opportunism.
Respondents in the Netherlands, a country characterized by high individualism, medium to
low scores on uncertainty avoidance, and low scores on masculinity and power distance, are more
likely to use considerate and creative voices and less likely to use patience and neglect. They are
37
indifferent toward exit, opportunism, and aggressive voice. The preference for active–
constructive responses and rejection of passive–destructive responses may be attributed to the
effect of high individualism and low power distance and masculinity. The low preference for
passive–destructive responses also implies that the effect of individualism is stronger than the
effect of power distance and masculinity on their response strategies.
Compared with respondents from the other countries, those in Switzerland, a country
characterized by high power distance, medium to high scores on individualism and uncertainty
avoidance, and medium scores on masculinity, are less likely to prefer exit, aggressive voice, and
patience but are indifferent toward opportunism, creative and considerate voices, and neglect.
The medium to high levels of individualism and uncertainty avoidance explain the low preference
for passive response strategies. However, the high level of power distance does not seem to
influence response strategy preference. The Swiss respondents revealed the lowest scores on
aggressive voice, which cannot be explained by the moderating effect of national culture.
We find that respondents in Turkey, a country characterized by high scores on uncertainty
avoidance and power distance, medium scores on masculinity, and low scores on individualism,
are more likely to use patience, neglect, and opportunism; less likely to use exit and creative and
considerate voices; and indifferent toward aggressive voice, compared with respondents from the
other countries. The preference for passive response and low preference for active strategies
stems from the low level of individualism. Although the high level of uncertainty avoidance
seems not to influence response strategy preference, the preference for destructive response
strategies is affected by a high score on power distance and medium score on masculinity. The
exit results cannot be explained by the direct of effect of cultural dimensions, but the relatively
low preference for exit, compared with that in other countries, may be explained by Turkish
respondents’ lower sensitivity to social satisfaction and alternative availability.
38
Finally, respondents in the United Kingdom, a country characterized by high scores on
individualism and masculinity and low scores on power distance and uncertainty avoidance, are
more likely to prefer exit and aggressive, creative, and considerate voices; less likely to use
patience and neglect; and indifferent toward opportunism. High individualism leads to a
preference for active but not for passive responses, whereas high masculinity results in a
preference for destructive responses.
Taken together, these results indicate that the four cultural dimensions, both directly and
interactively with exchange variables, affect the degree of activeness–passiveness and
constructiveness–destructiveness of response strategy preference. Specifically, individualism–
collectivism and uncertainty avoidance influence the degree of activeness–passiveness and
moderate the effect of economic satisfaction and alternative availability. In addition, masculinity–
femininity and power distance influence the degree of constructiveness–destructiveness and
moderate the effect of social satisfaction. The interaction between country and alliance-specific
investments is not significant. One explanation for this finding is that managers across cultures
perceive such investments as equally critical for relationship development, such that the
criticality of these investments overrides cultural differences.
6.2. Theoretical and managerial implications
Our study was conducted in an experimental setting with business students, however,
because the cultural effects we tested pertain to basic characteristics of human nature, we believe
that our results have some theoretical and managerial implications that are broader than the
narrow context in which they were tested.
Drawing on social exchange theory (Blau, 1964) and interdependence theory (Thibaut and
Kelley, 1959), response strategy research has proposed an investment model (e.g., Geyskens and
39
Steenkamp, 2000; Ping, 1993; Rusbult and Zembrodt, 1983) in which satisfaction with exchange
outcomes, specific investments, and the availability of alternatives drive preferences for response
strategies. In an international context, this model also should incorporate cross-cultural
differences, because people from different countries prefer different responses when confronted
with similar adverse situations. The results of our study contribute to response strategy literature
by demonstrating that preferences for response strategies vary across cultures.
Extant ISA studies demonstrate that cultural differences affect relationship development,
because they tend to exacerbate adverse situations (e.g., Barkema and Vermeulen, 1997;
Brouthers and Bamossy, 2006). We improve understanding of this issue by shifting attention
from the overall effect of cultural differences to the underlying direct and moderating
mechanisms through which national cultures influence managers’ decision making (Boyacigiller
and Adler, 1991). Just as people from different countries generally prefer different response
strategies, their reactions depend on different types of adversity, which corroborates the findings
of Thomas and Au (2002) and Thomas and Pekerti (2003). When faced with the same adverse
situation, our future managers respondents from different countries likely prefer different
response strategies, depending on the strategies they believe will be most adequate in their
cultural environment. This point is critical for the management of ISAs, because managers with a
better understanding of the behavioral intentions of their foreign partners, as well as their own
preferences, can better overcome adverse situations.
Prior studies (e.g., Brouthers and Bamossy, 2006; Meschi, 1997) also suggest that learning
about cultural differences can reduce the negative effect of adverse situations. By disentangling
the direct and moderating mechanisms through which national culture influences reactions to
adverse situations, our study provide managers with a better understanding of how to deal with
adverse situations in ISAs more effectively and thus should help them reduce the likelihood of
40
failure. An alliance manager from an individualistic country who understands that social
satisfaction may be more important than economic satisfaction for his or her counterpart from a
collectivist country will be better equipped to deal with the adverse situation and preserve the
alliance. Alliance managers who are able to predict which response strategies their foreign
counterparts prefer should be able to develop a collaborative relationship more easily and avoid
detrimental conflicts.
6.3. Limitations and further research directions
Although we believe our results offer important contributions to the international strategic
alliance literature and practice, our study is limited in several respects. First, we collected data in
only five countries; further studies should collect data from a wider variety of countries to further
disentangle the direct and moderating effects of national culture more clearly. Second, we
focused on the effect of national culture on response strategy preference, without measuring
cultural values directly; instead we used Hofstede’s scores to classify countries along the four
cultural dimensions. As these scores are not without critics and pertain to country level only,
measuring cultural values at the individual level would be necessary to understand both inter- and
intracountry differences. Third, by not measuring cultural values, we have been limited to testing
country differences rather than cultural differences. Therefore, our results should be interpreted
with caution, because other contextual variables, such as country wealth and level of
development, may have affected the results. Fourth, we conducted a scenario-based experiment
and measured behavioral intentions, which is appropriate for determining the direct and indirect
effects of national culture on response strategy preference. However, behavioral intentions do not
always translate into behavior. Fifth, the use of an experimental design with business students as
respondents increases the internal validity of the results but also may raise questions about
41
external validity. Although a comparison of the results of the Dutch student sample with those of
a survey of Dutch alliance managers indicates satisfactory level of convergence, our results about
the direct and moderating effects of culture might only be generalizable to a population of future
managers rather than actual managers. Therefore, further studies should conduct survey research
to validate the results with managers across countries. Studying managers’ work experience also
might be relevant, because our results indicate that age, which we use as a proxy for experience,
has a significant effect on two response strategies (i.e., creative and considerate voices).
7. Conclusion
In the context of response strategies, our results indicate that country differences directly
affect preferences for the activeness–passiveness and constructiveness–destructiveness of
responses. In addition, country differences appear to influence the perception of the adversity of
the situation. Although these results come from the particular context of international strategic
alliances, they provide valuable insights into the effect of national culture on response behavior in
general. Response strategies share similarities with coping strategies (e.g., Radford et al., 1993)
and conflict resolution styles (e.g., Morris et al., 1998), which also entail active–passive and
constructive–destructive behaviors. Thus, our findings about the differential effects of country
may apply in these contexts too, though this assertion requires empirical confirmation.
42
References
Aharoni, Y., Brock, D.M., 2010. International business research: Looking back and looking
forward. Journal of International Management 16, 5–15.
Ariño, A., de la Torre, J., Ring, P.S., 2001. Relational quality: Managing trust in corporate
alliances. California Management Review 44, 109–131.
Barkema, H.G., Vermeulen, F., 1997. What differences in the cultural backgrounds of partners
are detrimental for international joint ventures? Journal of International Business Studies 28,
845–864.
Bateman, T.S., Zeithaml, C.P., 1989. The psychological context of strategic decisions: A model
and convergent experimental findings. Strategic Management Journal 10, 59–74.
Bello, D., Leung, K., Radebaugh, L., Tung, R., Van Witteloostuijn, A. 2009. From the Editors:
Student samples in international business research. Journal of International Business Studies
40, 361–364.
Blau, P.M. 1964. Exchange and Power in Social Life. Wiley, New York.
Boyacigiller, N.A., Adler, N.J. 1991. The parochial dinosaur: Organizational science in a global
context. Academy of Management Review 16, 262–290.
Brouthers, K.D., Bamossy G.J., 2006. Post-formation processes in Eastern and Western European
joint ventures. Journal of Management Studies 43, 203–229.
Buchanan, L., 1992. Vertical trade relationships: the role of dependence and symmetry in
attaining organizational goals. Journal of Marketing Research 29, 65–75.
Byrne, B.M., 2001. Structural Equation Modeling with AMOS: Basic Concepts, Applications,
and Programming. Lawrence Erlbaum Associates, Mahway, NJ.
Croson, R., Anand, J., Agarwal, R. 2007. Using experiments in corporate strategy research.
European Management Review 4, 173–181.
Das, T.K., 2006. Strategic alliance temporalities and partner opportunism. British Journal of
Management 17, 1–21.
Doney, P.M., Cannon, J.P., Mullen, M.R., 1998. Understanding the influence of national culture
of the development of trust. Academy of Management Review 23, 601–620.
Farrell, D., 1983. Exit, voice, loyalty, and neglet as responses to job dissatisfaction: A
multidimensional-scaling study. Academy of Management Journal 26, 596–607.
Geyskens, I., Steenkamp, J.-B.E.M., 2000. Economic and social satisfaction: Measurement and
43
relevance to marketing channel relationships. Journal of Retailing 76, 11–32.
Hagedoorn, M., Yperen, N.W., Van de Vliert, E., Buunk, B.P., 1999. Employees’ reactions to
problematic events: A circumplex structure of five categories of respones, and the role of job
satisfaction. Journal of Organizational Behavior 20, 309–321.
Hibbard, J.D., Kumar, N., Stern, L.W., 2001. Examining the impact of destructive acts in
marketing channel relationships. Journal of Marketing Research 38, 45–61.
Hirschman, A.O., 1970. Exit, Voice and Loyalty: Responses to Decline in Firms, Organizations
and States. Harvard University Press, Cambridge, MA.
Hofstede, G., 2001. Culture’s Consequences, 2nd ed. Sage, Thousand Oaks, CA.
Hofstede, G., Usunier, J.-C., 2003. Hofstede’s dimensions of culture and their influence on
international business negotiations, in Ghauri, P.N., Usunier, J.-C. (Eds.), International
Business Negotiation, 2nd ed. Elsevier, Oxford, pp. 137-153.
Hu, L.T., Bentler, P.M., 1999. Cutoff criteria for fit indexes in covariance structure analysis:
conventional criteria versus new alternatives. Structural Equation Modeling 6, 1–55.
Huberty, C. J., Morris, J. D., 1989. Multivariate analysis versus multiple univariate analyses.
Psychological Bulletin 105, 302–308.
Jiang, X., Li, Y., Gao, S., 2008. The stability of strategic alliances: Characteristics, factors, and
stages. Journal of International Management 14, 173–189.
Joardar, A., Kostova, T., Ravlin, E.C., 2007. An experimental study of the acceptance of a
foreign newcomer into a workgroup. Journal of International Management 13, 513–537.
Johnson, J. L., Cullen, J. B., Sakano, T., Takenouchi, H. 1996. Setting the stage for trust and
strategic integration in Japanese-U.S. cooperative alliances. Journal of International Business
Studies, 27, 981–1004.
Kirkman, B.L., Lowe, K.B., Gibson, C.B. 2006. A quarter century of “Culture’s Consequences”:
A review of empirical research incorporating Hofstede’s cultural values framework. Journal of
International Business Studies 37, 285–320.
Leck, J. D., Saunders, D. M. 1992. Hirschman’s loyalty: Attitude or behavior? Employee
Responsibilities and Rights Journal, 5, 219–230.
Lee, J., Jablin, F.M., 1992. A cross-cultural investigation of exit, voice, loyalty and neglect as
responses to dissatisfying job conditions. Journal of Business Communication 29, 203–228.
McDonald, R.P., Ho, M.-H.R., 2002. Principle and practice in reporting structural equation
44
analyses. Psychology Methods 7, 64–82.
Meschi, P.-X., 1997. Longevity and cultural differences of international joint ventures: toward
time-based cultural management. Human Relations 50, 211–228.
Morris, M.W. et al., 1998. Conflict management style: Accounting for cross-national differences.
Journal of International Business Studies 29, 729–748.
Nakana, C. 1971. Japanese society. Berkeley: University of California Press.
Nakos, G., Brouthers, K.D., 2008. International alliance commitment and performance of small
and medium-size enterprises: The mediating role of process control. Journal of International
Management 14, 124–137.
Nunnally, J.C., Bernstein, I.H., 1994. Psychometric Theory, 3rd ed. McGraw-Hill, New York.
Park, S.H., Ungson, G.R., 2001. Interfirm rivalry and managerial complexity: A conceptual
framework of alliance failure. Organization Science 12, 37–53.
Perrinjaquet, A., Furrer, O. Usunier, J.-C., Cestre, G., Valette-Florence, P. 2007. A test of the
quasi-circumplex structure of human values. Journal of Research in Personality 41, 820–840.
Ping, R.A., 1993. The effects of satisfaction and structural constraints on retailer exiting, voice,
loyalty, opportunism, and neglect. Journal of Retailing 69, 320–352.
Pressey, A.D., Qu, X.X., 2007. Buyer-supplier relationship dissolution: The Chinese context.
Journal of Business & Industrial Marketing 69, 320–352.
Radford, M.H.B., Mann, L., Ohta, Y., Nakane, Y., 1993. Differences between Australian and
Japanese students in decisional self-esteem, decisional stress, and coping styles. Journal of
Cross-Cultural Psychology 24, 284–297.
Ralston, D.A., Egri, C. P., Reynaud, E., Srinivasan, N., Furrer, O., et al. in press. A 21st century
assessment of values across the global workforce. Journal of Business Ethics, forthcoming.
Rusbult, C.E., Farrell, D. 1983. A longitudinal test of the investment model: The impact on jobsatisfaction, job commitment, and turnover of variations in rewards, costs, alternatives, and
investments. Journal of Applied Psychology 68, 429–438.
Rusbult, C.E., Farrell, D., Rogers, G., Mainous, A.G., 1988. Impact of exchange variables on
exit, voice, loyalty, and neglect: An integrative model of responses to declining job
satisfaction. Academy of Management Journal 31, 599–627.
Rusbult, C.E., Zembrodt, I.M., 1983. Responses to dissatisfaction in romantic involvements: A
multidimensional-scaling analysis. Journal of Experimental Social Psychology 19, 274–293.
45
Rusbult, C. E., Zembrodt, I. M., Gunn, L. K. 1982. Exit, voice, loyalty, and neglect: responses to
dissatisfaction in romantic involvements. Journal of Personality and Social Psychology, 43,
1230–1242.
Steenkamp, J.-B.E.M., Baumgartner, H., 1998. Assessing measurement invariance in crossnational consumer research. Journal of Consumer Research 25, 78–90.
Strahan, R., Gerbasi, K.C., 1972. Short, homogeneous versions of Marlow-Crowne social
desirability scale. Journal of Clinical Psychology, 28, 191–193.
Taras, V., Rowney, J., Steel, P., 2009. Half a century of measuring culture: Review of
approaches, challenges, and limitations based on the analysis of 121 instruments for
quantifying culture. Journal of International Management 15, 357–373.
Taras, V., Steel, P., Kirkman, B.L. in press. Improving national cultural indices. A meta-analysis
of Hofstede’s dimensions. Journal of World Business, forthcoming.
Thibaut, J.W., Kelley, H.H. 1959. The Social Psychology of Groups. Wiley, New York.
Thomas, D.C., Au, K., 2002. The effect of cultural differences on behavioral responses to low job
satisfaction. Journal of International Business Studies 33, 309–326.
Thomas, D.C., Au, K., Ravlin, E.C., 2003. Cultural variation and the psychological contract.
Journal of Organizational Behavior 25, 451–471.
Thomas, D.C., Pekerti, A.A., 2003. Effect of culture on situational determinants of exchange
behavior in organizations: A comparison of New Zealand and Indonesia. Journal of CrossCultural Psychology 34, 269–281.
Tjemkes, B., Furrer, O. 2010. The antecedents of response strategies in strategic alliances.
Management Decision 48, 1103–1133.
Triandis, H.C., 1995. Individualism and Collectivism. Westview Press, Boulder, CO.
Wathne, K.H., Heide, J.B., 2000. Opportunism in interfirm relationships: Forms, outcomes, and
solutions. Journal of Marketing 64, 36–51.
Withey, M.J., Cooper, W.H., 1989. Predicting exit, voice, loyalty, and neglect. Administrative
Science Quaterly 34, 521–539.
Yum, Y.-O., 2004. Culture and self-construal as predictors of responses to accommodative
dilemmas in dating relationships. Journal of Social and Personal Relationships 21, 817–835.
Zhou, J., George, J.M., 2001. When job dissatisfaction leads to creativity: Encouraging the
expression of voice. Academy of Management Journal 44, 682–697.
46
Table 1. Scores of Hofstede’s Four Cultural Dimensions by Country
Country
Individualism–
Collectivism
Masculinity–
Femininity
Power
Distance
Uncertainty
Avoidance
Japan
46
M
95
H
54
M
92
H
The Netherlands
80
H
14
L
38
L
53
M-L
Switzerland (Frenchspeaking part)
64
M-H
58
M
70
H
70
M-H
Turkey
37
L
45
M
66
H
85
H
United Kingdom
89
H
66
H
35
L
35
L
Mean
51
51
51
64
Notes: H = high, M = medium, L = low.
Source: Hofstede, 2001.
47
A form of voice that consists of efforts to win
without consideration for the concerns of the
exchange partner (Hagedoorn et al., 1999).
A form of voice that consists of the generation
of novel and potentially useful ideas (Zhou
and George, 2001).
A form of voice that consists of attempts to
solve the situation by considering one’s own
concerns as well as those of the partner (Ping,
1993).
Abiding relationship issues in silence with the
confidence that things will get better
(Hagedoorn et al., 1999; Hibbard et al., 2001;
Ping, 1993)
Passively allowing the relationship to
deteriorate (Ping, 1993; Rusbult et al., 1982)
Aggressive
voice
Creative
voice
Considerate
voice
Patience
Neglect
I do not plan anything extra to solve the situation with the XXX alliance.
I will not initiate anything to improve the situation with the XXX alliance.
I will not deal with the situation with the XXX alliance.
I trust the situation will solve itself.
I will not say anything to XXX because I expect the situation to go away by itself.
I will wait patiently and hope the situation solves itself.
I will try to come to an understanding with XXX to solve the situation.
I will approach XXX with a proposition to solve the situation and work it out together.
I will contact XXX to find an acceptable solution for both partners.
I will adopt a fresh approach to improve the situation.
I will come up with new ideas to improve the situation with the XXX alliance.
I will propose creative solutions to XXX to improve the situation.
I will prove in all possible ways to XXX that my firm's solution for the situation is right.
I will be very persistent with XXX to have them accept my firm's solution to the situation.
I will strongly advocate my firm’s solution to solve the situation with the XXX alliance.
I will purposefully exaggerate the situation in order to get additional benefits.
I will deliberately make the situation sound more problematic than it really is to obtain more
benefits from the XXX alliance.
I will withhold important information from XXX to gain additional benefits.
I think that I will probably stop doing business with XXX.
I am not likely to continue the alliance with XXX.
I believe that I will terminate the XXX alliance.
The disinclination to continue the current
relationship (Ping, 1993; Rusbult et al., 1982;
Withey and Cooper, 1989)
Self-interested behavior that is explicitly or
implicitly prohibited (Ping, 1993; Wathne and
Heide, 2000).
Sample Item
Description
Opportunism
Response
Strategies
Exit
Table 2. Response Strategies, Descriptions, and Sample Items
48
d.f.
168
168
168
167
168
921
χ2
256.9
251.4
259.0
329.9
207.2
1 440.0
.00
.00
.00
.00
.00
.02
p-value
1.56
1.57
1.50
1.54
1.98
1.23
.020 [.018, .022]
.047 [.036, .057]
.039 [.028, .048]
.046 [.035, .057]
.059 [.050, .069]
.031 [.013, .044]
.049
.058
.049
.052
.059
.046
χ2/d.f. RMSEA [90% CI] SRMR
.94
.95
.96
.95
.90
.95
NNFI
A constraint is placed between two error terms, reducing the degrees of freedom to 167.
**
Four factor loadings and nine intercept constraints are released to achieve partial scalar invariance (∆χ2 = 44.5, p = .11, ∆CFI = .001).
*
Model
Country Models
Japan (n = 262)
The Netherlands (n = 334)
Switzerland (n = 255)
Turkey (n = 278)*
U.K. (n = 250)
Multigroup Model (n = 1,379)
Partial scalar invariance model**
Table 3. Overall Fit Indices of the Measurement Models
.95
.96
.97
.96
.92
.96
CFI
49
Table 4. Construct Reliability and Partial Correlation Matrices
A. Japan
CR AVE
.86
.68
.74
.49
.60
.34
.85
.68
.75
.51
.76
.51
.80
.58
Response Strategy
Mean
1 Exit
3.57
2 Opportunism
3.01
3 Aggressive voice
4.48
4 Creative voice
5.35
5 Considerate voice 5.22
6 Patience
2.31
7 Neglect
2.25
n = 262.
s.d.
1.44
1.23
1.00
1.07
1.11
1.10
1.10
α
.86
.74
.60
.84
.75
.76
.80
1
2
.25***
.15*
-.25***
-.48***
.05
.25***
.31***
-.26***
-.28***
.38***
.42***
Response Strategy
Mean
1 Exit
3.09
2 Opportunism
3.48
3 Aggressive voice
4.54
4 Creative voice
5.56
5 Considerate voice 5.54
6 Patience
1.92
7 Neglect
2.14
n = 334.
s.d.
1.40
1.19
1.06
.86
.87
.82
.87
α
.89
.70
.74
.81
.75
.74
.71
B. The Netherlands
CR AVE
1
.89
.73
.71
.45
.25***
.74
.49
.20***
.82
.61
-.30***
.75
.49
-.47***
.75
.50
-.07
.71
.45
.22***
Response Strategy
Mean
1 Exit
2.95
2 Opportunism
3.28
3 Aggressive voice
4.11
4 Creative voice
5.54
5 Considerate voice 5.37
6 Patience
1.92
7 Neglect
2.22
n = 255.
s.d.
1.44
1.24
1.09
.88
.99
.92
.99
α
.89
.72
.72
.78
.72
.77
.77
C. Switzerland
CR AVE
1
.89
.73
.74
.49
.06
.72
.47
.09
.80
.59
-.42***
.72
.46
-.47***
.77
.53
.01
.77
.53
.34***
Response Strategy
Mean
1 Exit
3.13
2 Opportunism
3.85
3 Aggressive voice
4.40
4 Creative voice
5.27
5 Considerate voice 5.15
6 Patience
2.32
7 Neglect
2.58
n = 278.
s.d.
1.31
1.22
1.04
1.06
1.02
1.17
1.17
α
.82
.67
.68
.78
.68
.78
.71
D. Turkey
CR AVE
.83
.61
.69
.43
.68
.41
.81
.59
.68
.42
.78
.54
.71
.45
2
.48***
-.07
-.11*
.04
.12*
2
.41***
-.06
-.16**
.21***
.35***
1
2
.10†
.03
-.40***
-.48***
.23***
.48***
.32***
.08
-.04
.09
.11
3
.13*
.04
.01
-.01
3
.06
-.01
-.12*
-.01
3
.08
.01
-.02
.08
3
4
5
6
.60***
-.35*** -.22***
-.45*** -.43*** .69***
4
5
6
.55***
-.20*** -.10†
-.37*** -.32*** .46***
4
5
6
.60***
-.27*** -.24***
-.47*** -.50*** .51***
4
5
6
.26***
.15*
.72***
†
-.11
-.37*** -.30***
-.06
-.57*** -.55*** .59***
E. United Kingdom
Response Strategy
Mean s.d.
α
CR AVE
1
2
3
4
5
6
1 Exit
3.76
.80
.68
.69
.43
2 Opportunism
3.76
.68
.69
.69
.43
.07
3 Aggressive voice
4.78
.69
.77
.77
.53
.12† -.01
4 Creative voice
5.70
.65
.68
.68
.41
.06
-.02
.14*
5 Considerate voice 5.50
.74
.64
.64
.38
.03
.03
.01
.25***
†
6 Patience
1.94
.65
.77
.78
.54
.12
-.02
-.02
-.28*** -.16*
7 Neglect
2.20
.69
.64
.68
.44
-.02
.04
-.01
-.21** -.10
.38***
n = 250.
Notes: s.d. = standard deviation. α = Cronbach’s alpha. CR = composite reliability. AVE = average variance
extracted.
†
p < .10; *p < .05; **p < .01; ***p < .001.
50
F
F
F
F
F
F
F
F
F
F
F
F
Economic satisfaction
Social satisfaction
Alliance sp. investments
Alternatives available
Country × Econ. Sat.
Country × Soc. Sat.
Country × Alliance sp. invest.
Country × Alt. available
Age
Gender
Problem severity
Social desirability
***
***
Aggressive
Voice
***
Creative Voice
***
Considerate
Voice
Patience
***
Neglect
***
.25
2.81†
.07
24.87***
5.15***
14.24***
1.04
2.47*
1.67
.98
.40
5.71*
.02
.91
.39
.22
3.87*
2.64
1.87
3.62**
2.34†
4.71***
36.82***
67.20***
5.84*
92.31***
3.42***
2.03
.05
1.70
.54
.88
1.71
.42
.16
1.54
8.07**
3.83*
.24
6.89***
5.67*
.69
.02
16.33***
4.07*
.25
3.52†
8.06**
4.33***
3.88**
1.31
.48
2.91*
4.03**
2.10†
.38
2.46*
1.11
1.48
1.34
1.48
4.56***
.99
.04
10.80***
1.93
10.53***
11.93***
.15
4.54*
.73
10.01**
11.78***
60.77***
.01
2.71†
1.93
32.11***
2.60***
.72
.01
.12
3.95*
.75
1.11
.73
2.40*
.02
.77
2.45
8.77**
22.53
20.27
15.79
11.96
8.74
11.09
9.04
(UK = JA) ≥ NL TU > (UK = UK > (NL = JP UK > (SW = NL) (UK = NL) ≥ SW (TU = JP) > (SW TU > (JP = SW
≥ (TU = SW) NL = SW) > JP = TU) > SW
≥ (JP = TU)
≥ (JP = TU)
= UK = NL)
= UK = NL)
***
Opportunism
.99*
.99
.98**
.97***
.96***
.96**
.98
.96**
(2.13)
(.70)
(3.17)
(5.69)
(2.20)
(1.81)
(1.08)
(2.06)
.96*** (7.93)
.93*** (13.76)
.99**
(2.65)
.92*** (16.80)
.77*** (12.96)
Wilks Λ (F)
51
Notes: Covariates appearing in the model are evaluated with the following values: Age = 23.6 years; Gender = 1.59; Problem severity = 4.23; Social desirability = 5.23. L = low, H =
High.
1
Pairwise comparison significant at 5%.
†
p < .10; * p < .05; ** p < .01; *** p < .001.
Corrected Model F-Value
F
Country
Pairwise comparision1
Exit
Table 5. MANCOVA Results
Figure 1. Interaction Effects between Exchange Variables and Countries
Japan
The Netherlands
Switzerland
Turkey
United Kingdom
52
APPENDIX: SCENARIO TEXTS
General Introduction
Imagine that you are an alliance manager at a firm called BIOPHARM. As an alliance manager you are responsible for all of the
strategic alliances BIOPHARM is engaged in. It is your job to assure that the performance of each alliance meets BIOPHARM’s
expectations. You have the authority and power to make any necessary decisions concerning the future of these alliances. In
BIOPHARM’s portfolio of alliances, one of the alliances is an international alliance with a foreign partner, STARTECH. This
alliance was established five years ago and is equally important for both companies.
Manipulations Texts
Economic satisfaction
Positive
During the past year, the benefits BIOPHARM derived through the
STARTECH alliance met the firm’s expectations. The financial
objectives BIOPHARM had set were fully attained. More
specifically, the alliance generated more revenues for BIOPHARM
than it had initially expected.
Negative
During the past year BIOPHARM’s benefits from the STARTECH
alliance did not meet the firm’s expectations. The financial
objectives BIOPHARM had set were not fully attained. More
specifically, the alliance generated fewer revenues for BIOPHARM
than it had initially expected.
Social satisfaction
Positive
Up until now, STARTECH has been very cooperative and flexible
when making necessary adjustments to the alliance. For example,
necessary renegotiations to change contractual clauses were usually
settled quickly. Consequently, the working relationship with
STARTECH went very smoothly as both partners trusted one
another completely and BIOPHARM did not need to allocate
additional resources to monitor the alliance.
Negative
Up until now, STARTECH has not been very cooperative and
flexible when making necessary adjustments to the alliance. For
example, on one occasion it took extensive renegotiations to change
some minor contractual clauses. Consequently, the working
relationship with STARTECH became more troublesome since both
partners started to distrust one another and BIOPHARM had to
allocate additional resources to monitor the alliance.
Alliance specific investments
Positive
In the past, BIOPHARM made some minor reusable investments to
make the alliance with STARTECH work. For example,
BIOPHARM has invested into a database, which is useful with or
without the STARTECH alliance. In addition, only few additional
costs are to be expected if BIOPHARM should terminate the
STARTECH alliance. For example, no penalty fee would have to be
paid to STARTECH if BIOPHARM prematurely ends the alliance.
Negative
In the past, BIOPHARM made substantial specialized investments
that would be lost if the STARTECH alliance were to be terminated.
For example, BIOPHARM has invested into a tailor-made database,
which is only useful, if exploited together with STARTECH. In
addition, substantial costs can be expected if BIOPHARM should
end the STARTECH alliance. For example, a high penalty fee has to
be paid to STARTECH if BIOPHARM prematurely ends the
alliance.
Alternatives availability
Positive
At this moment, BIOPHARM has plentiful alternatives available.
For example, launching a new product without STARTECH can be
considered as a serious alternative. Moreover, developing new
products without STARTECH’s help is possible with only a few
additional investments. In addition, various other firms can be
viewed as potential partners to replace STARTECH and these firms
are interested in forming an alliance with BIOPHARM.
Negative
At this moment, BIOPHARM has few serious alternatives available.
For example, launching a new product without STARTECH is too
risky. Moreover, developing new products without STARTECH’s
help is too costly, as substantial investments would be required. In
addition, only a few other firms can be viewed as potential partners
to replace STARTECH. In addition, most of these firms are already
engaged in other alliances with BIOPHARM’s competitors.
53